July 3, 2026
Saved time, missing money
AI saves about 3% of your hours, and almost none of it reaches the money
AI steals an hour of your week, but commenters say don’t expect a bigger paycheck
TLDR: A major study found AI saves workers about an hour a week, but almost none of that turns into higher pay or fewer hours. Commenters immediately split into two loud camps: people saying “obviously” and people insisting the research is already outdated because AI is changing too fast.
The big plot twist in this Danish workplace study is brutally simple: AI really does save time—about 2.8% of work hours, roughly an hour a week—but almost none of that shows up in people’s pay. Researchers matched surveys from 25,000 workers with actual payroll data, and the result was a lot less sexy than the flashy “AI makes you 40% faster” headlines. In real jobs, across real weeks, the gains are small, messy, and apparently very easy for companies to just… absorb.
And the comment section? Absolute food fight. One camp basically said, “Well, duh,” with pdp arguing that saving time was never guaranteed to mean more money, comparing AI to ads and backlinks: expensive, unavoidable, and not always profitable. Then came the anti-study squad. killerstorm called it “a weird kind of bullshit,” while lazzlazzlazz said it was “just embarrassing” to believe research like this right now. Translation: some readers think the economists are measuring yesterday’s AI in a world where the tools change every five minutes.
The funniest drive-by came from chapel, who claimed the article itself looked AI-written and tapped out immediately: “That’ll be a no from me dawg.” Meanwhile peter422 dropped the ultimate chaos grenade: last year AI wrote 0% of their code, now it writes 100%. So the community verdict is deliciously split: AI is either a mild time-saver with no cash prize, or the study is already obsolete because the robots are moving faster than the researchers can hit publish.
Key Points
- •A Danish study linking AI-adoption surveys with payroll records found average time savings of about 2.8% of work hours across roughly 25,000 workers in 7,000 workplaces.
- •The article says the Danish evidence showed no significant impact from chatbots on earnings or recorded hours across occupations.
- •The article contrasts small real-world weekly gains with much larger task-level improvements reported in controlled AI studies.
- •A cited experiment found ChatGPT reduced time spent on certain professional writing tasks by 40% while improving graded quality by 18%.
- •A Harvard-BCG field experiment cited in the article found GPT-4 improved performance on tasks within its range but reduced correctness on a task set outside that range.